Fit a model predicting energy consumed per capita (energy) from the percentage of residents living in metropolitan areas (metro) in the States.xlsx file. Use Excel and Word for your responses as necessary.
Excel cannot perform regression analysis on words, so if you use “Region” in your model, make sure to create a dummy column for it, specifically add a new column and specially code the word “South” as a “1,” the term “West” as a “2,” etc. Then run the regression using the new column of coded values.
Generate the following as at least part of your analysis.
- Show the table(s) showing the Excel regression analysis printout(s). 10 pts.
- From your analysis, show how to compute the average energy consumed. 5 pts.
- Include a Spearman Test analysis if the best model has only one x value . 5 pts.
- Your recommendation as to whether the model is statistically significant and use the Causal Test to recommend if there is a cause and effect relationship (or the degree of cause and effect) with some other related model. Show your reference used for the Causal Test (literature, a phone call to an authority in the field, or some other reference). Use plain English for your explanation. 10 pts.
Now, run a stepwise regression on this same data set to find the “best” model for energy consumption.
1. Develop and test several regression models, and identify which model seems to be best. 10 pts
2. Justify why it is the best model, based on its associated statistics. 5 pts.
3. Develop an estimated regression equation that can be used to justify the salary level. 5 pts.
4. Discuss how to use the causal test to complete the analytics. 5 pts.
0 comments